DKVP --incr-key option (#1839)

* Code support for --incr-key

* Add source code for online help for new flag

* Run `make dev`

pkg/cli/option_parse.go

@@ -104,6 +104,7 @@ var FLAG_TABLE = FlagTable{
 		&CSVTSVOnlyFlagSection,
 		&JSONOnlyFlagSection,
 		&PPRINTOnlyFlagSection,
+		&DKVPOnlyFlagSection,
 		&CompressedDataFlagSection,
 		&CommentsInDataFlagSection,
 		&OutputColorizationFlagSection,
@@ -523,6 +524,31 @@ var PPRINTOnlyFlagSection = FlagSection{
 	},
 }
 
+// ================================================================
+// DKVP-ONLY FLAGS
+
+func DKVPOnlyPrintInfo() {
+	fmt.Println("These are flags which are applicable to DKVP format.")
+}
+
+func init() { DKVPOnlyFlagSection.Sort() }
+
+var DKVPOnlyFlagSection = FlagSection{
+	name:        "DKVP-only flags",
+	infoPrinter: DKVPOnlyPrintInfo,
+	flags: []Flag{
+
+		{
+			name: "--incr-key",
+			help: "Without this option, keyless DKVP fields are keyed by field number.  For example: `a=10,b=20,30,d=40,50` is ingested as `$a=10,$b=20,$3=30,$d=40,$5=50`.  With this option, they're keyed by a running counter of keyless fields.  For example: `a=10,b=20,30,d=40,50` is ingested as `$a=10,$b=20,$1=30,$d=40,$2=50`.",
+			parser: func(args []string, argc int, pargi *int, options *TOptions) {
+				options.WriterOptions.BarredPprintOutput = true
+				*pargi += 1
+			},
+		},
+	},
+}
+
 // ================================================================
 // LEGACY FLAGS
 

pkg/cli/option_types.go

@@ -53,11 +53,12 @@ type TReaderOptions struct {
 	irsWasSpecified            bool
 	allowRepeatIFSWasSpecified bool
 
-	UseImplicitHeader   bool
-	AllowRaggedCSVInput bool
-	CSVLazyQuotes       bool
-	CSVTrimLeadingSpace bool
-	BarredPprintInput   bool
+	UseImplicitHeader    bool
+	AllowRaggedCSVInput  bool
+	CSVLazyQuotes        bool
+	CSVTrimLeadingSpace  bool
+	BarredPprintInput    bool
+	IncrementImplicitKey bool
 
 	CommentHandling TCommentHandling
 	CommentString   string

pkg/input/record_reader_dkvp_nidx.go

@@ -14,7 +14,7 @@ import (
 	"github.com/johnkerl/miller/v6/pkg/types"
 )
 
-// splitter_DKVP_NIDX is a function type for the one bit of code differing
+// line_splitter_DKVP_NIDX is a function type for the one bit of code differing
 // between the DKVP reader and the NIDX reader, namely, how it splits lines.
 type line_splitter_DKVP_NIDX func(reader *RecordReaderDKVPNIDX, line string) (*mlrval.Mlrmap, error)
 
@@ -169,25 +169,42 @@ func recordFromDKVPLine(reader *RecordReaderDKVPNIDX, line string) (*mlrval.Mlrm
 
 	pairs := reader.fieldSplitter.Split(line)
 
+	// Without --incr-key:
+	//   echo 'a,z=b,c' | mlr cat gives 1=a,z=b,3=c
+	//   I.e. implicit keys are taken from the 1-up field counter.
+	// With it:
+	//   echo 'a,z=b,c' | mlr cat gives 1=a,z=b,2=c
+	//   I.e. implicit keys are taken from a 1-up count of fields lacking explicit keys.
+	incr_key := 0
+
 	for i, pair := range pairs {
 		kv := reader.pairSplitter.Split(pair)
 
 		if len(kv) == 0 || (len(kv) == 1 && kv[0] == "") {
 			// Ignore. This is expected when splitting with repeated IFS.
 		} else if len(kv) == 1 {
-			// E.g the pair has no equals sign: "a" rather than "a=1" or
+			// E.g. the pair has no equals sign: "a" rather than "a=1" or
 			// "a=".  Here we use the positional index as the key. This way
 			// DKVP is a generalization of NIDX.
-			key := strconv.Itoa(i + 1) // Miller userspace indices are 1-up
+			//
+			// Also: recall that Miller userspace indices are 1-up.
+			var int_key int
+			if reader.readerOptions.IncrementImplicitKey {
+				int_key = incr_key
+			} else {
+				int_key = i
+			}
+			str_key := strconv.Itoa(int_key + 1)
+			incr_key++
 			value := mlrval.FromDeferredType(kv[0])
-			_, err := record.PutReferenceMaybeDedupe(key, value, dedupeFieldNames)
+			_, err := record.PutReferenceMaybeDedupe(str_key, value, dedupeFieldNames)
 			if err != nil {
 				return nil, err
 			}
 		} else {
-			key := kv[0]
+			str_key := kv[0]
 			value := mlrval.FromDeferredType(kv[1])
-			_, err := record.PutReferenceMaybeDedupe(key, value, dedupeFieldNames)
+			_, err := record.PutReferenceMaybeDedupe(str_key, value, dedupeFieldNames)
 			if err != nil {
 				return nil, err
 			}
@@ -204,9 +221,9 @@ func recordFromNIDXLine(reader *RecordReaderDKVPNIDX, line string) (*mlrval.Mlrm
 	var i int = 0
 	for _, value := range values {
 		i++
-		key := strconv.Itoa(i)
+		str_key := strconv.Itoa(i)
 		mval := mlrval.FromDeferredType(value)
-		record.PutReference(key, mval)
+		record.PutReference(str_key, mval)
 	}
 	return record, nil
 }